Nvidia’s Jetson Thor Powers the Dawn of General Robotics Era

Nvidia has launched the Jetson AGX Thor developer kit and production modules, a next-generation robotics platform designed to power millions of robots with advanced AI capabilities. The Jetson AGX Thor is powered by the NVIDIA Blackwell GPU and features 128GB of memory, delivering up to 2,070 FP4 teraflops of AI compute within a 130-watt power envelope. This represents a significant performance leap compared to its predecessor, the Jetson Orin, with 7.5x higher AI compute and 3.5x better energy efficiency. These enhancements enable real-time reasoning and inference, critical for highly performant physical AI applications such as humanoid robotics, surgical assistance, and autonomous industrial machines [1].

The Jetson AGX Thor developer kit is now available for $3,499, and production modules are available through distribution partners. The platform is already attracting industry leaders, including Agility Robotics, Amazon Robotics, Boston Dynamics, Caterpillar, Figure, Hexagon, Medtronic, and Meta, as early adopters. Companies such as John Deere, OpenAI, and Physical Intelligence are evaluating the platform for their physical AI projects. According to Jensen Huang, CEO of NVIDIA, the Jetson AGX Thor is “the ultimate supercomputer to drive the age of physical AI and general robotics,” enabling developers to run multiple generative AI models at the edge [1].

The Jetson AGX Thor is designed to address one of the most significant challenges in robotics: enabling robots to interact in real-time with the physical world. The platform supports a wide range of generative AI models, including vision language action (VLA) models and large language and vision language models. It also runs the full NVIDIA AI software stack, including NVIDIA Isaac for robotics, NVIDIA Metropolis for video analytics AI agents, and NVIDIA Holoscan for sensor processing. This integration allows developers to build and deploy applications such as visual AI agents that monitor worker safety, humanoid robots capable of complex manipulation tasks, and smart operating rooms that guide surgeons using multi-camera data streams [1].

Early adopters have highlighted the potential of Jetson AGX Thor to transform robotics in their respective industries. Agility Robotics, a leader in humanoid robotics, plans to adopt Jetson Thor as the onboard compute platform for the sixth generation of its robot, Digit. This transition is expected to enhance Digit’s real-time perception and decision-making capabilities, supporting more complex AI skills and behaviors. Similarly, Boston Dynamics is integrating Jetson Thor into its Atlas humanoid robot, enabling it to leverage server-level compute and high-bandwidth data processing for more advanced AI workloads [1].

Beyond humanoids, Jetson AGX Thor will accelerate various robotic applications, including surgical assistants, smart tractors, delivery robots, industrial manipulators, and visual AI agents. The platform’s real-time inference capabilities on larger, more complex AI models open new possibilities for multimodal AI applications. For instance, research labs at Stanford University, Carnegie Mellon University, and the University of Zurich are using Jetson Thor to push the boundaries of perception, planning, and navigation models for potential applications in medical triage, search and rescue, and unstructured environments. With the Jetson ecosystem supporting thousands of developers and 150+ hardware and software partners, the platform is well-positioned to accelerate innovation in robotics and AI [1].

The Jetson AGX Thor developer kit is available now, and production modules are available for purchase through authorized NVIDIA partners. The platform’s integration with the NVIDIA Jetson software stack and its compatibility with cloud-to-edge AI workflows ensure a seamless development experience for developers. As the demand for physical AI grows across industries, the Jetson AGX Thor is expected to play a pivotal role in enabling the next generation of intelligent, autonomous robots that can operate in complex, dynamic environments [1].

Source:

[1] NVIDIA Blackwell-Powered Jetson Thor Now Available ... (https://nvidianews.nvidia.com/news/nvidia-blackwell-powered-jetson-thor-now-available-accelerating-the-age-of-general-robotics)

[2] title2 (url2)

[3] title3 (url3)